860 ALASKAN GLACIER STUDIES
On the other hand, the Ripon Glacier is not in exactly the right position for this, and the narrowness of the shoal area is difficult to explain by glacial erosion and easy to understand if the whole deposit is a submerged moraine. The coincidence of a halt of the glacier exactly on an uneroded rock swell is also difficult to explain, if the moraine is thought of as resting so precisely on an irregularity in rock. There is no special rock texture or structure to help explain this phenomenon, if mainly a rock feature. The geological map by Grant and Higgins l indicates that Willard Island and the whole shore line of Blackstone Bay are made up of graywacke and slate and small dikes with no observed peculiarities, such as would fit in with the idea of differential erosion due to the nature of the rock. On the whole, therefore, we believe that the facts point toward a moraine bar here, rather than a narrow upward swell in the bedrock floor of the fiord.
Relationships of Forest to Stages of Qlaciation. As noted by Grant and Higgins, the southern six-sevenths of Willard Island constitutes a relatively barren zone, though not one of the extremely young barren zones like those near Ripon and Lawrence Glaciers, or the ones near Barry and other glaciers in Harriman Fiord, for within it there are a few trees up to 10 or 12 inches in diameter. Throughout this somewhat overgrown barren zone, there is a scattered growth of evergreen trees and many alder thickets, but the larger part of the area has no trees, though in the open spaces thickly covered with grasses, moss, and peat. On the north it terminates along a definite line (marked as the limit of the barren zone, b, Fig. 53), extending westward across Willard Island from the base of the morainic bar. North of this line there is thick, mature forest, and it therefore seems clear that an advance of the former glacier of Blackstone Bay extended down to this line some scores of years and possibly over a century ago. The fact that the edge of this barren zone continues the line of the moraine bar just described, suggests the association of the moraine with this advance. It is to be noted, however, that we found no similar moraine at the western end of this zone, in the channel west of Willard Island; though the fact that our soundings there were a quarter of a mile apart, makes it possible that there is a narrow submerged moraine which we did not discover.
All about lower Tebenkof Glacier and throughout the northern end of Blackstone Bay, there is thick, mature spruce forest extending from timber line to sea level. The forest ends about a mile north of Carroll Glacier on the western side of the bay and at Ripon Glacier on the eastern side; but there is scattered vegetation beyond this. Badger Point, for example, has scattered trees and thick groves with individuals apparently as old as those on the southern part of Willard Island; and at moderate heights above sea level there are a few clumps of good-sized spruces even on the mountain spur between Beloit and Blackstone Glaciers. The trees nearest the glaciers, as on Badger Point and between Beloit and Blackstone Glaciers and near Tebenkof Glacier, are tall and well formed, but some of the trees in the southern part of Willard Island are much stunted, very thick in proportion to height, and with limbs only on the northern side, showing that in exposed places the glacier wind, here from the south, has retarded tree development. Even with the limitations, resembling the conditions in southern Harriman Fiord, which of course lengthen the time since great glacial expansion, the abrupt limit of mature, thick forest on Willard Island, coinciding, as it does,
i Bull. 443, U. S. Geol. Survey. 1910, PI. H